Premature infants are at high risk of suboptimal health and development. These risks greatly increase when prematurity is combined with 2 or more social-environmental risks, such as poverty, minority status, and low maternal education. This randomized clinical trial will test the impacts of a developmentally based intervention, H-HOPE (Hospital-home transition: optimizing prematures'environment), for infants born between 29-34 weeks gestational age (GA) with at least two social-environmental risk factors. This target group has the biologic capability to benefit from intervention. H-HOPE is innovative because it integrates two components used successfully in prior research but never before combined. Infant remediation using a developmentally appropriate multisensory intervention addresses the specific behavioral organization needs of premature infants. Maternal redefinition and re-education by a nurse-community advocate team uses participatory guidance to address the needs of mothers of premature infants. The synergistic effect of these simultaneous improvements for infant and mother should lead to: 1) more mature infant behavioral organization and hospital progression;2) improved maternal recognition of infant behavioral cues, greater confidence in infant care, more positive perception of the infant, and lower anxiety;3) more positive mother- infant interaction and greater mother-infant contingency;4) improved infant development and growth;and 5) lower infant health care utilization and costs. H-HOPE provides intervention from 32 weeks GA to one month corrected age, a time of transition to oral feeding, from the hospital to home, and from hospital to outpatient providers, when mothers of premature infants express need for support. We will randomly assign 252 infants to the H-HOPE or the Attention Control group. Power analysis shows that with an 80 percent retention rate, we will have adequate power to identify expected intervention effects. Variables are measured during hospital stay, at intake, immediately prior to discharge, and at six weeks corrected age. Analyses employ Hierarchical Linear Modeling clustered within clinical sites, with infant sex, biologic and social-environmental risk factors as covariates. If successful, H-HOPE will provide a national model for improving early infant health and development and reducing health costs. For example, reducing hospital stays by just three days for the almost 500,000 infants born prematurely could save over two billion dollars annually.